Retrieving Twitter argumentation with corpus queries and discourse analysis
We propose a corpus linguistic approach for retrieving argumentation from tweets about Brexit. We analyse two corpora, one from before the referendum in 2016 and one from early 2019. Our approach is based on the manual development of morphosyntactic corpus queries (in CQP syntax) that target specific argumentation patterns. For the present corpora we have developed 130 queries targeting 34 logical formulae. The paper showcases the query development and presents quantitative and qualitative results on how Brexit-related arguments have changed on Twitter from 2016 to 2019.
Article outline
- 1.Introduction
- 2.Related work
- 2.1Everyday argumentation
- 2.2Corpus linguistics and argumentation
- 3.Data basis
- 3.1Data acquisition and corpus preparation
- 3.2Corpus statistics
- 3.3Parts of speech and phrases
- 3.4Lemmas and entities
- 4.Methodology
- 4.1Corpus queries for logical representation
- 4.2Query development
- 5.Results
- 5.1Quantitative overview
- 5.2Case Study: As NP I VP
- 6.Discussion
- 7.Conclusion
-
Notes
-
References